Insane Hydraulics
Bold and audacious blog about fluid power
This post comes from one of those shop situations when I realize that it could have been an awesome video only after I have something fixed, and shooting anything is all too late now. Ma-a-a-n... Still, I do have a couple of pictures and a couple of hydraulic tips to share, so I'll give it my best try (and leave a mental note to shoot a video the next time I come across a similar failure).
So, today we will be testing and fixing a relief valve, more exactly - a pilot-operated spool-type relief valve from this little baby here:
This is a hydraulically piloted two-section directional control valve from a Komatsu front loader. It is large and intimidating but is actually a very simple (and very classic) DCV that has a two-pump inlet that by-passes one of the pumps to the tank after a set pressure is reached (for torque-limiting purposes), an electric override for this system (for light work-mode operation), the main relief valve, a couple of anti-shocks and anti-cavitations and that's it.
It came to our shop because the owner thought it was stalling the engine, which turned out not to be the case, but still, when I get a valve on a test bench, I like to check all the settings (because that's how any bench test should be approached). And as I started checking the relief valve, I immediately saw that it was "acting out".
I definitely should have realized at that very moment that it was the perfect opportunity to make a video about a "properly malfunctioning" relief valve, and then about a "proper" way to test it, and then a "proper" way to fix it (possibly, of course), but, as I said, I did just the opposite - I fixed it first, and only then realized how great and educational a video it would have made...
Let me begin with the test, then. Verbally. The set-up is simple - I have a pilot pressure source to pilot the DCV spools, one pump feeding the main P port, a second pump feeding the secondary P inlet, a load-inducing valve in a work port, and, of course, there are pressure gauges in the P lines and also a flow meter in the T (part of the test bench).
Then I loosen the main relief valve pressure setting screw, pilot the DCV, close the loading valve, and as I rotate the setting screw - I look at the pressure gauge (an analog pressure gauge, by the way, because in that case seeing the pressure rate change is more important than the very exact pressure value), and see how well my input is mirrored by the pressure change.
When I check (set) relief valves, I usually do it in "half-turn in, half-turn out, full-turn in" increments. It is a good way to notice the valve's response and the rate of the pressure change to turns relation. Ideally, the gauge needle should faithfully follow your input. And any change in the valve's pressure response will immediately "feel strange".
That was exactly what I saw - I saw the pressure steadily climb up to about 160 bar, and then the climb slowed and then stopped and became unstable - a sure sign that something was not OK.
So, here's another tip. When something like this happens to a piloted relief valve, it can mean two things - either something else is opening to tank somewhere (like a missing seal, an improperly set anti-shock valve, or a damaged actuator) or a damaged pilot section in the relief valve itself. In that case, if you are testing a DCV on a test bench that has its own relief valve protecting your pressure source, what you can do is keep turning the adjusting screw in till the spring bottoms out and locks the pilot poppet. Usually, you will see a sharp pressure increase at that point, and this will mean that indeed the pilot section (and not an external leak) is the culprit.
A warning now - use this tip very, very wisely! First - never do this on an actual machine. Blocking an outlet of a fixed displacement pump is not a fun experience, believe me. And of course, be aware of the fact that not all valves can do that. For example not all pilot arrangements will allow for bottoming out the adjusting spring and locking the pilot cone. And make sure that your test bench has the relief valve set appropriately!
When you know what you are doing, this is actually very simple - you see the pressure act out, then you turn the crew a couple of turns it, feel the resistance of it bottoming out and see the pressure climb abruptly - and you know immediately that the pilot section is busted. Either a seal is missing or the cone of the pilot poppet has a nasty groove in it. Bam! Five seconds and your diagnostics is done!
That was exactly the case. I bottomed the screw till I felt it starting to lock, and I saw the pressure jump up. Great - I know now that it is the relief valve's pilot and not a mysterious leak somewhere else in the valve ( which, it could have easily been by the way). So, I remove the relief valve cartridge, disassemble it, and here's what I find:
A classic conical poppet with classic wear. And now let us proceed to yet another shop trick (so to speak) to bring this poppet back to life.
Since this part is tiny, cutting this wear down in a lathe is not easy, because the tooling marks will be hard to polish out. Sanding paper and high speed could do the job, but when the wear is deep it takes a lot of time. A much faster way is to use a high-speed rotary tool and a stone point. Even this "on the knee" arrangement works wonders:
As you can see, I have a 1/4'' pneumatic grinder strapped to a lathe carrier. Talk about "redneck engineering"! Still, here's the finished product (a finishing touch was done with fine grit sanding paper):
Doesn't look too bad, does it? So, I put it all back together again and re-tested the vale, and I worked flawlessly this time!
No video (for now), but I still leave you with three shop tips for today: